High-Stakes Digital CSI

May 23, 2016Articles New Jersey Law Journal

Matthew S. Adams and Jordan Kaplan authored the New Jersey Law Journal article, "High-Stakes Digital CSI."

Much ink has been spilled addressing the privacy and security arguments surrounding Apple's showdown with the U.S. Department of Justice (DOJ) earlier this year in connection with the massive law enforcement response and investigation into the most deadly terrorist attack on U.S. soil since Sept. 11, 2001. This is not another examination of those heavily debated legal issues. Rather, drawing on the facts that have been revealed publicly about the San Bernardino investigation, we set out here to present a case study on what can go wrong when digital evidence is mishandled, and to present a series of best practices for handling ESI—an acronym for "electronically stored information" that has been known to evoke dread and angst in even the most seasoned legal and investigative teams.

On Dec. 2, 2015, 14 people were killed and 22 were seriously injured at the Inland Regional Center in San Bernardino, California, in a despicable, abhorrent act of terrorism. The attack was carried out by Syed Rizwan Farook and Tashfeen Malik, a married couple that apparently had become radicalized jihadists somewhere along the way, despite otherwise conducting themselves as ordinary residents of a California suburb. The pair carried out their attack during a training event and holiday party for the San Bernardino County Department of Public Health. Farook was an employee of the San Bernardino County Department of Public Health, fueling speculation that his actions were that of a disgruntled employee that had lost his mind. Americans collectively watched as initial reports suggested another mass shooting in yet another soon-to-be infamous corner of the country. Soon, however, our dread quickly got worse when an even more terrifying plot emerged, and the reality set in that a likely ISIS-inspired terror attack had occurred on American soil. That night, Farook and Malik were killed in a shoot-out with police, who confronted the pair on a California street. With their deaths, any chance that their motives and affiliations with larger plots would be uncovered through interrogation was lost. It would come down to the tedious investigative task of reconstructing the pair's entire lives, including their use of cellphones, computers and other digital media that have become the hallmarks of modern society.

That painstakingly laborious task unfolded largely out of the spotlight at first. Then, on Feb. 16, a U.S. magistrate judge sitting in the Central District of California entered an order requiring Apple, a purely private actor with no known involvement in the FBI's ongoing investigation other than to have designed, manufactured and sold the iPhone assigned to Farook by his employer, to assist the DOJ's investigation by overriding security features baked into the technology that drives the device occupying nearly 50 percent of the U.S. marketplace for smartphones. Apple, led by CEO Tim Cook, immediately took to an aggressive public relations campaign vehemently opposing the government's efforts, while Apple's lawyers sought to overturn the magistrate judge's Order in court. The story occupied headlines for nearly six weeks until, in late March, the DOJ announced that it had unlocked Farook's iPhone without Apple's help, and abruptly abandoned its quest to compel Apple's assistance with the investigation.

Like many issues with national security implications, one magistrate judge's ruling in California reverberated throughout Washington, D.C. In the midst of the legal battle between Apple and the DOJ, FBI Director James Comey testified before members of Congress regarding the San Bernardino investigation. Director Comey was forced to acknowledge that the FBI made significant missteps in how it handled Farook's iPhone in the investigation. Most specifically, Director Comey testified that when the shooter's iPhone was initially recovered, the FBI requested that the owner of the phone—Farook's employer, San Bernardino County—reset the password to the iCloud storage account linked to the phone. "I do think, as I understand from the experts, there was a mistake made in the … 24 hours after the attack, where the county at the FBI's request, … took steps that made it hard—impossible later to cause the phone to backup again to the iCloud." As a consequence of that early error in the investigation, a forensic examination of Farook's actual physical device became necessary for the most up to date version of the data contained thereon, setting off the very public debate that followed.

No person or agency is infallible. Setting aside the debate over privacy versus security that has flowed from the events of Dec. 2, 2015, there can be no question that the FBI's investigative intentions in the case were pure. Nonetheless, the FBI's acknowledged missteps in the San Bernardino terror case serve as a stark reminder that basic fluency in the collection of digital evidence is a prerequisite to practicing law in 2016. There are some very straightforward rules of the road that must always be followed when it comes to the collection of digital evidence that is to be used in a legal proceeding:

(1) Know the mechanics of the media upon which the evidence resides. The DOJ would not have been forced into the position it was with Apple had it thought through the precise way that data was stored on Farook's iPhone, and where duplicate copies of that data might reside. For lawyers, this does not mean obtaining a degree in computer science. Instead, it means using the skills of lawyering (questioning, deductive reasoning and critical analysis) to assemble the people most knowledgeable of the mechanics of the subject media to provide the necessary details. Where digital evidence can be secured from multiple sources, it should be secured from all of them to paint the most complete picture. Redundancy is never a bad thing when it comes to digital forensics, as backups frequently uncover data that was deleted in a more localized setting. With ESI, industry standard de-duplication processes can easily eliminate the burden of multiple copies of the same thing by comparing the unique digital identifiers assigned to each discrete file, like its MD-5 hash value.

(2) Learn how your media was secured and how it has been used.Just as the FBI no doubt learned early on that Farook's iPhone and related cloud storage accounts were actually the property of his employer, implicating a host of time-saving Fourth Amendment consent to search issues, an awareness of how media was secured and how it has been used can be the difference between a dump of meaningless data from a digital forensics analysis and finding the smoking gun. For example, one of the primary ways that data theft occurs is by use of portable storage media, like external hard drives and USB drives. So-called "link file analysis"—a means of creating a correlation between the use of such portable devices on a computer and access to certain data residing either on that computer or a server based upon forensically extracted time line data—is a powerful way to prove the where, when, how and sometimes who, pertaining to a breach. Yet, discrete digital media standing alone, without details on how they were used and procured, would not provide the details necessary to conduct such a useful correlative analysis.

(3) Prepare and rehearse a collection plan.The well-known axiom "measure twice, cut once" is very fitting here. The benefit of a table-top data collection exercise before an actual collection attempt is made cannot be overstated. A rehearsal will often point out flaws in your plan, or can identify weaknesses and vulnerabilities with the subject data. Even if certain variables cannot be overcome, knowing what they will be when it is time for the actual collection does not always come from abstractly pondering the plan alone. Live simulation tends to work out or at least identify the kinks before they impede your collection efforts.

(4) Execute the collection plan under forensically acceptable conditions.Collecting digital information is not like making a photocopy, unless, of course, forensically acceptable methods are employed. The simple reason for this is that digital evidence contains metadata, or descriptive data about the data. ESI is a multidimensional, layered composite of discoverable information, the totality of which is relevant and necessary to a complete understanding of the weight, scope and overall import of the ESI as a piece of evidence in the case. Without the multidimensional composite, the ESI lacks reliability, may not be capable of authentication, and, because of the manner in which ESI is maintained, is easily manipulated, corrupted and/or deleted. Simply stated, a one-dimensional snapshot of ESI is worse than a photocopy of only one side of a double-sided paper document. Because lawyers are not, for the most part, digital forensics experts, they are not accustomed to operating under forensically acceptable methods such as, for example, the use of a write block device when copying digital media to avoid alteration of metadata. This, coupled with the likely need for testimony to authenticate and lay the foundation for the evidence, invariably means that a digital forensics expert must be employed to collect the data. Beware, "dragging and dropping" files, copying and pasting information and saving open files onto a new drive or device do not constitute forensically acceptable conditions. The main problem with nonforensic data collection is that important metadata is overwritten and replaced, which could play a vital role in the over-arching investigation that has led to the data collection in the first place.

(5) Preserve originals and only conduct analysis from duplicates. There are frequently many layers of review necessary to a complete understanding of what, if any, role ESI will play. Some analysis, especially analysis by lawyers who frequently do not have the type of complex software packages that allow data to be reviewed without alteration of the underlying source data readily at their disposal, may modify aspects of ESI like metadata inadvertently. For example, a review of a USB drive's worth of Microsoft Excel spreadsheets will corrupt the underlying data if that review is conducted by a lawyer simply plugging the drive into his or her computer and exploring its contents. At the very least, if such a review is to take place, the review should be done from a forensically copied image of the source drive that is specifically designated as a "Review Copy Only" to make clear that its contents have not been altered or modified, and it is being used exclusively for informational purposes only while the original copy and any other forensic copies of the original are safely stashed away in a secure location.

Most lawyers will never be forced to reveal a data collection blunder to Congress the way that FBI Director Comey had to when discussing the San Bernardino terror attack investigation. Yet, as reflected in most modern frameworks for the collection and production of ESI, such as the protocols developed by the Joint Working Group on Electronic Technology in the Criminal Justice System (JETWG), which state as "Principal One" that "[l]awyers have a responsibility to have an adequate understanding of electronic discovery[,]" it is apparent that attorneys must have at least a basic grasp on electronic discovery—from collection to production—in order to carry out their ethical responsibilities to their clients. Accordingly, the basic principles for ESI collection discussed herein are a necessary starting point considering the abundance of ESI that has become such a ubiquitous component of our daily lives.

Reprinted with permission from the May 23 issue of New Jersey Law Journal. (c) 2016 ALM Media Properties, LLC. Further duplication without permission is prohibited. All rights reserved.